prismane16/monitor-accuracy-under-rl download history
prismane16/monitor-accuracy-under-rl is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 499 times (499 in the last 7 days), and 499 times in total. It ranks #33,178 among datasets by monthly downloads.
Cached artifacts for "What Happens to a Monitor's Accuracy When You Train Against It" Abraham Yeung and Anagha Ramaswamy, Stanford University. Interpretability as a Science (InterpScience) Workshop, NeurIPS 2026. Code and paper: https://github.com/Abraham-y/reader-writer-probe-rl Thes